The most important functions of the shared resource are to establish and maintain databases that enable access to clinical data for research purposes. The shared resource also supports other informatics efforts that enhance the ability of the Consortium to conduct clinical research in an environment increasingly sensitive to human subjects' protections and regulatory accountability. For 30 years, transplant data collection and distribution at Fred Hutchinson Cancer Research Center (FHCRC) has been a centralized activity, with a comprehensive database now of over 12,000 transplants performed at FHCRC that is used routinely by all Consortium members engaged in clinical transplant research. These data are housed in an Ingres database referred to as Gateway. Data from approximately 400 new transplants are added each year and over 3,000 patients are being followed as survivors of transplant. This provides a large database for characterizing the short- and long-term effects of transplant on survival and quality of life. The shared resource comprises dedicated data collection staff responsible for abstracting and entering data from the entire transplant treatment course and a Transplant Informatics Group (TIG) that develops and maintains the Gateway database. The Solid Tumor Informatics Group (STIG) was established in 2002 to develop and operate the computing infrastructure needed to support the solid tumor clinical research programs. Research databases are currently available for prostate and breast cancer subjects; future databases will include gastroenterology, gynecology, and lung/thoracic subjects. This group is responsible for the design, development, enhancement, implementation, and oversight of the Consortium Oncology Data Integration (CODI) data warehouse that aggregates clinical data for solid tumor research. Data enters the warehouse from automated data feeds and web-based data entry systems based on a system called Caisis. Although Caisis originated from an application specifically designed to support a prostate cancer database, it is adaptable to other disease populations. At present the prostate and breast cancer databases are populated with data from approximately 3,500 and more than 16,000 subjects, respectively.